Abstracts/Lung Cancer 14 (19%) 377-408 319

studies of cancer risk. The validity of measurements of DNA adducts formal from hydrophobic aromatic hydrocarbons in peripheral blood mononuclear cells (MNCs) was investigated by comparing the levels of aromatic DNA adducts detected in lung tissue from 31 lung cancer patients with those detected in MNCs from the same individuals using the )2P-postlabeling assay. The associations of smoking history and intake of dietary antioxidants with adduce levels also were assessed Tissue- specific, as well as common DNA adduces were detected in lung and blood; total MNC adduce levels were highly correlated with total lung adduces. After smoking cessation, adduct levels appeared to decay inboth tissuesat similar rates. Multivariateanalyses(Poissonregression modeling) indicated that dietary antioxidant intake (carotenoids, vitamin A, and retinal) modified the levels ofaromatic DNA adduces in both the lungs and blood. Of all models tested, the optimal one for predicting lung adduce levels included the measure of blood MNC adduct levels only. Therefore, blood MNCs area valid surrogate tissue for estimating the burden of DNA adduces in respiratory tissue in molecular epidemiological studies.

Radon and risk of cancer: Epidemiological studies after occupational and domestic exposure Tirmarche M. IPSN, Dept. Proteecrion Sante Homme/Dosim., Service d’Evaluation/Gestion Risques, B.P. No. 6. F-92265 Fontenay aux Roses Ceder. Rev Epidemiol Sante Pub1 1995;43:451-60.

The evaluation of cancer risk after exposure to radon is mainly based on the results of uranium miners follow-up. A cohort study on the French uranium miners has demonstrated an excess of lung cancer and of larynx cancer mortality. A linear dose-response relationship has been described between the excess relative risk of lung cancer and the cumulative exposure to radon (poisson regression). This shrdy has contributed to a joint analysis of 1 I cohorts of miners, the aim being a more precise evaluation of the different factors able to influence the dose-response relationship between radon and lung cancer mortality. These factors are: age at first exposure, attained age, time since exposure, the pattern of exposure over time and tobacco consumption. The extrapolation of the risk for the general public from the risk estimated after occupational exposure, has to be considered by taking in account several remarks: uranium miners are exposed, beside radon, to hvo other radiological components, gamma rays and long lived uranium dust, and to other substances specific of the mines, which are absent in the domestic environment but may with radon have an effect on the lung cancer risk. It was impossible to estimate directly, from these uranium miners data, the risk linked to radon for non-smokers and for female population. A case control-study is currently be carrying out in the French hospitals, in order to estimate the risk of lung cancer linked to the last 30 years of radon exposure in the dwellings.

Potential role df environmental and domestic exposure to tremolite in pleural cancer in New Caledonia Goldberg P, Lute D, BillowGalland MA, Quenel P, Salomon-Nekiriai C, Nicolau J et al. iNSERb Unite&?, Hopital National de St. -Maurice, 14 Rue du Vald ‘Owe, F-94415 Sainte-Maurice Ceder. Rev Epidemiol Sante Pub1 1995;43:444-50.

A previous study of respiratory cancers in New Caledonia (197% 1987) showed an excess risk of pleural cancer in this South Pacific French Territory, leading to the identification of an environmental pollution. In some villages, the residents use for their houses a white- wash made from a rock derived from local outcroppings. Analysis of samples of rock and whitewash showed that they consisted of tremolite asbestos. High levels of tremolite were detected in airborne samples collected in these villages and in biological specimens of patients with pulmonary cancer or mesothelioma; the concentrations of fibers are up

to 78000 fibers per litre of air and 44 millions of fibers per gramme of dry tissue. Besides the whitewash, the environmental exposure to tremolitefiberscouldalsobeassociatedwithcertain occupations. A case control study under process will allow the estimation of respiratory cancer risks associated with the exposure to tremolite.

Industries and occupations at increased risk of lung cancer mortality in Turin (1981-89) and Italy (1981-82) Lagorio S, Forastiere F, Rapiti E, Di Pietro A, Costa G. Istituto Superiore di Sanita, Laboratorio Igiene Ambientale. Viale Regina Eltwa, 299, 00161 Roma. Med Lav 1995;86:309-24.

In the framework of an occupational disease surveillance program, based on integration of current information systems, the first Italian occupational mortality study was carried out. This paper reports on excess lung cancer risk by industry and occupation. The study population consists of subjects included in the Italian Cross-Sectional Study (STI) and in the Turin Longitudinal Study (SLT), both of which are surveys based on record-linkage procedures between census records and death certificates. The ST1 is a six-month follow-up of Italian residents at the 1981 census. The SLT is a prospective study of Turin residents at the 198 1 census, followed for mortality up to 1989. Only persons aged 18- 64 years at entry, and economically actives, were eligible for the occupational mortality analysis (i.e. 15,734 deceased individuals out of 13 million subjects in the STI, and 435,608 individuals, among whom 10,789 deaths OUXOT~~, in the SLT). Information about job and economic activity recorded at census consisted of the Italian standard 1981 industry and occupation codes. Lung cancer relative risks by category of industry and job were estimated as mortality odds ratios (MOR) in the STI, and as observed to expected death ratios (SMR) in the SLT. Only excess risks based on 3 observed cases and with p < 0.1, were included in the present report. Lung cancer mortality was increased in different industries and jobs. The excess risks found in the mechanical and transport industries are of particular interest in a public health perspective, due to the high number of Italian workers employed in these sectors. From an etiological point of view, however, careful attention should be paid to the excess lung cancer risks among workers in the wood manufacturing industry, in meat preparation, and in nursing occupations, where detailed analytical studies of exposure profile and cancer risk are warranted.

Risk of smoking for squamous and small cell carcinomas of the lung modulated by combinations of CYPlAI and GSTMI gene polymorphisms in a Japanese population Kihara M, Kihara M, Noda K. Department of Epidemiology, Cancer CenterResearch Institute. Yokohama. Carcinogenesis 1995; 16:2331-6.

Genes for cytochrome P4501Al (CYPIAI) and glutathione S- transferwe class mu (GSTM 1) have been shown to be polymorphic, and have been implicated in tobacco-related carcinogenesis. In the present study, the role of the combined genotypes CYPlAl and GSTMl as a possible modulator of smoking related lung cancers was studied in relation to the tobacco smoke exposure level in 118 Japanese patients aged < 70 with squamousor small cell carcinomas of the lung. Among male smoking patients, the overall proportion of the GSTMI null genotype (GSTMl[-1) was slightly higher than among healthy male smoker controls (56.7 % versus48.146, P = 0.17). Littledifferencewas observed between smoker patients and corresponding controls in overall frequencies of m2 mutant allele homozygotes (CYPlAl[m2/m21) (16- 18 %) and Val encoding allele homozygotes (5-6 %). However, when subjects were categorized by both CYPlAl genotype (Mspl polymorphism) and GSTMl genotype, GSTMl(-) became markedly more expressed in patients with CYPIAl(m2/m2) when compared to the corresponding smoker controls (81.3% versus 39.446, P < 0.01).

380 Abstracts/Lung Cancer 14 (1996) 377-408

When odds ratioswereestimatedusing nonsmoking patients and healthy coqtrols as a reference, the relative risk for developing lung cancer was found to increase in a cigarette dose-dependent manner across all combinations of genotypes. Furthermore, a 7- to 8-fold variation in risk was found among the various combinations; 3.2 in individuals with combined GSTMl(+) and CYPlAl(nQlm2) and 21.9 in those with combinedGSTM I(-)andCYPlAl(niXn2)genotypewhen thesmoking index (6cigarettes smoked per day x years of smoking) was set at 800. The results suggest that individuals having CYPlAl(nQlm2) are relatively resistant to tobacco-related lung cancers when combined with GSTMl( +), but are highly susceptible when combined with GSTMl(- ). Combined CYPlAl and GSTMl genotype is thus a Potential predictor of genetic susceptibility to smoking-related lung cancers in populations where CYPlAl m2 or Val alleles are common.

Me&analysis of studies of lung cancer among silicotics Smith AH, Lopipero PA, Barroga VR. Department of Envtl. Health Sciences, School of Public Health, University of California. Berkeley. CA 94720. Epidemiology 1995;6:617-24.

The relation between exposure to crystalline silica and lung cancer

has been a controversial topic, and findings haveappeared inconsistent. In this paper, we focus on lung cancer risks in epidemiologic studies of silicotics. Weabstracteddatafrom29shtdiesforquantitativeevaluation. We identified several studies that suffered from biases due to competing risks of different causes of death - in particular, death due to silicosis itself. After adjustment for competing risks, all 29 studies demonstrated lungcancerrelativerisk(RR) estimatesgreaterthanone. ThepooledRR estimate for the 23 studies that could be combined was 2.2, with a 95 96 confidence interval (CI) of 2.1-2.4. The pooled estimates by study design were 2.0 (95 % CI = 1.8- 2.3) for cohort studies and 2.5 (95 % CI = 1.8-3.3) for case-control studies. The proportional mortality studies combined gave a summary RR of 2.0 (95% CI = 1.7-2.4), whereas the studies of cancer incidence gave a summary RR of 2.7 (95 % CI = 2.3-3.2). Although statistical tests demonstrated heterogeneity between studies, and the confidenceintervals given abovemay therefore be a little too narrow, the overall tindigs could not be attributed to chance, confounding by smoking, or other sources of bias. We conclude that the association between silicosis and lung cancer is causal, either due to silicosis itself, or due to adirect effect of the underlying exposure to silica.

An ecological study of diet and lung cancer in the South Pacific Le Marchand L. Hankin J.H. Bach F. Kolonel L.N. Wilkens L.R. Stacewin-Sapuntzakis M. et al. Epidemiology Program, CancerResear& Center of Hawaii, University of Hawaii, 1236 Lauhala Street, Hon~lu- lu. HI 96813. Int J Cancer 1995;63:18-23.

Incidence rates of lung cancer have been markedly lower for Fiji than for other South Pacific countries, despite similar rates of smoking. We conducted population-based surveys in several island nations of the South Pacific (Cook Islands, Fiji Tahiti and New Caledonia) and used data from Caucasian Japanese, Hawaiian, Filipino and Chinese controls in a case-control study of lung cancer in Hawaii to investigate the role of diet in explaining differences in lung cancer incidence among 20 ethnic-sex groups. In a stepwise linear regression of lung cancer rates on smoking, diet and other variables, smoking, as expected, explained the majority (61%) of the variability in incidence. However, several dietary components also explained significant portions of the variance. Lutein intake explained 14% and vitamin E intake, cholesterol intake and height explained 5-7 % each of the remaining variance in incidence. Associations with lutein and vitamin E were inverse, whereas those with cholesterol and height were direct. Dietary O-carotene intake was not associated with lung cancer incidence. These ecological data provide

evidence for a protective effect of lutein against lung cancer. A protective effect of dietary vitamin E and a risk-enhancing effect of dietary cholesterol are also suggested.

Basic biology

Properties of classic protein kinase C in human small cell lung carcinoma NCLH345 cells Jones CLA, Beck LK, Broma JP, Holley M, Dempsey EJ, Kane MA. Denver Veterans Affair Medical Ctr., 1055 Clement Street, Denver, CO 80220. Cell Growth Differ 1995;6: 1627-34.

Bombesin-like peptides (BLPs) activate protein kinase C (PKC), which leads to proliferation in nonmalignant Swiss 3T3 cells. The purpose of this study was to determine if PKC expression in the classic human small cell lung carcinoma NCI-H345 cell line, which has an autocrine growth loop involving BLPs, exhibited unique properties that could result in malignant behavior. PKC activity and phorbol dibutyrate binding in NCI-H345 cells had properties similar to other reports. PKC activity in the cytosolic fraction Increased to 100% as cells proliferated through lag, log, and plateau growth states. However, during the first 3 days after plating (lag growth state), 40-50% of the PKC activity was membrane associated, indicating a substantial portion in an activated form, possibly a result of BLP autocrine stimulation. NCl-H345 cells expressed the PKC isoenxymes 6,&D, zeta and r but not gamma or -, a pattern different from Swiss 3T3 cells or normal brain. Further characterization of the Ca*‘/phospholipid-dependent (classic) PKC iso- enzymes, 6 and 8, showed that PKC 8 was predominantly cytosolic (80%) as expected, but PKC 6 was primarily membraneassociated (SO- 90 46). Exposure of NCI-H345 cells to 200 nM phorbol 12-myristyl 13- acetate rapidly (within 2 min) decreased cytosolic PKC activity, with no change in the particulate activity, but did not alter [‘HI-thymidine incorporation or subcellular distribution ofPKC 6or 8 by Western blot. These results suggest altered PKC regulation in human small cell lung carcinoma NCI-H345 cells, which could contribute to their malignant behavior.

Multiple chromosomal aberrations and llp allelotyping in lung cancer cell lines Bepler G, Koehler A. Med. /Molecular Cancer Biology Dept., Duke University Medical Center, Box 2610. Durham, NC 27710. Cancer Genet Cytogenet 1995;84:39-45.

Cytogenetic and molecular genetic studies have implicated many chromosomal aberrations in the pathogenesis of lung cancer. Deletions on 3p and 9p are presently the primary target for positional cloning of putative tumor suppressor genes. We have recently reported frequent loss of heterozygosity in three separate regions (HRAS, DllS12, Dl lSl6) on llp in freshly resected lung cancer specimens. Here we report cytogenetic and molecular genetic analyses of 26 permanently growing human lung cancer cell lines. Deletions indicating regions which may harbor potential tumor suppressor genes were found in 519 cell lines on 2p, S/9 on 2q, 619 on 3p, 719 on 3q, S/9 on 6q, 319 oo 9p, 5/9 on 1 lp, and 6/9 on 13q. Reduction to hemizygosity or a statistically significant increase in the frequency of homozygosity on 1 lp was found for all markers investigated except for ST5 (Dl lS832E). Eight of twenty-six (31 W) cell lines were hemixygous for Dl lS12 and 9/26 (35%) for DllS16. Seventeen of eighteen (94%) cell lines were homozygous for PTH (expected homozygosity, 53%), 15115 (100%) for WTl (expected homozygosity, 55X), and 16/18 (89%) for CAT (expected homoxygosity, 50%). These results confirm the notion that 1 Ip harbors several putative tumor suppressor genes which may become inactivated at different stages of tumor development and progression. They also provide a basis for selecting cell lines for genetic complementation specifically targeted at the regions described.